TRANSACTIONS OF SECTION F. — PKESIDENTIAL ADDBESS. 629 



Section F.— ECONOMIC SCIENCE AND STATISTICS, 

 President of the Section. — A. L. Bowley, M,A. 



THURSDA r, A UG UST 2. 



The President delivered the following Address : — 



From 1835 to 1855 Section F of the British Association was devoted to 

 ' Statistics/ and it is only from 1856 onwards that it has received its curious 

 name, ' Economic Science and Statistics.' It is interesting to recall thatBabbage 

 was its first President, and that in its earlier carepr such well-known pioneers in 

 the application of statistical method to industrial phenomena as Porter and Tooke 

 occupied the chair. In its later course economics and statistics have shared the 

 honour with public administration, whether related to trade or finance ; and in 

 recent years the professorial economist has alternated with the official adminis- 

 trator. It may be hoped that a new category will soon be added to the interesting 

 and varied list — that of those engaged in practical or applied economics, the 

 organisers of the army of industry; and in this connection it is much to be 

 regretted that Sir George Gibb was unable to take the place which I now occupy. 

 With a list which includes the names of Baines, Newmarch, Chadwick, Jevons, 

 Booth, Giffen, and Edgewortb, no complaint can be made that statistical science 

 and statistical art have not been worthily represented, and it would seem that 

 there was no species of the exponents of our group of sciences not already scheduled 

 in this roll ; but I find that I have the unique position of being the first professed 

 — or, shall I say, armchair .^ — statistician, with few economic credentials, to hold 

 this position ; and this fact leads me to direct my address mainly to the claims 

 of statistics to be an exact science, worthy to rank as such with those which 

 form the subject-matter of Sections A to L. Since, however, the title of our 

 section and the names of its officers, past and present, suggest the essential con- 

 nection of economics with statistics, and the establishment of both on a quadruple 

 basis of theory, history, experiment, and practice, my intention is to show that 

 our work resembles the natural sciences in the respect that the most delicate 

 researches in theory lead directly to visible and important pi-actical results. The 

 graduation of the income-tax, the supply of fish in the North Sea, and the expendi- 

 ture of a labourer's wage, are among the subjects to which I have recently had to 

 apply mathematical statistical analysis. 



It is a long step from Artliur Young's tours to Professor Edgeworths 

 * generalised law of great numbers,' but there is no distinction in the nature of 

 things between arithmetical and mathematical statistics,; the distinction to he 

 made is not between the various methods of accumulating and tabulating data, 

 but between the truth and falsity of the reasoning based on the tabulation. 

 Mathematical treatment in the end only furnishes us witli a microscope to observe 

 4ifferences which are blurrgd to the naked eye of arithmetic, and with a method 



